A literature review suggests that the onset of schizophrenic symptoms after the age of 45 is not rare. Yet, there have been few systematic studies of neuropsychological, brain-imaging and treatment-response characteristics of late-onset schizophrenia. We postulate that schizophrenia with onset after 45 is a heterogeneous entity with different subtypes. Some of these subtypes may be identifiable with certain clinical, neuropsychological and brain morphological evaluations, and may be associated with differences in neuroleptic response. Our 6- month pilot study of 14 late-onset schizophrenics shows both the feasibility of and the need for the proposed investigation. We will study 90 DSM-III-R late-onset schizophrenics over a period of 5 years. Patients will be assessed with selected psychiatric, neurologic and neuropsychological measures at baseline and then systematically followed at four-month intervals. MRI scans of the brain will be done at the time of study-entry, and will be analyzed with quantitative densitometric and volumetric methods. Normal controls matched for age, gender, level of education and socioeconomic status will be evaluated and followed in a manner similar to late-onset schizophrenics. We will evaluate neuroleptic response in the late-onset schizophrenic patients in terms of therapeutic benefit (comparing "drug-free" psychopathology ratings with those after 6 weeks of haloperidol), and risk of tardive dyskinesia (comparing TD ratings at the end of the study). On the basis of the literature review and our data, we predict that one subset of patients diagnosed as having late-onset schizophrenia will have significant neuropsychological deficits and structural abnormalities in MRI, and poor therapeutic response to neuroleptics with greater risk of tardive dyskinesia. A small proportion of such patients will be found, at follow-up, to have a diagnosable dementing disorder, that initially presented with a schizophrenia-like clinical picture. Another subset of patients will be similar on various measures to age-matched normal controls. We believe that our findings will help improve the understanding of the diagnosis, neurobiological subtyping and neuroleptic treatment of late-onset schizophrenia.